ClearSet.AI

We hope this new year thus far is bringing you excitement of what’s to come. Enjoy your day and focus on the future! Leave 2024 behind because today is a new day and a new year! 🎊🥳 The team at ClearSet.AI says to you all… HAPPY NEW YEAR! 🎆🎈Let’s welcome in 2025! 🤖🦾🥂 #2025 #newyears #happynewyear #artificialintelligence #aisoftwaredevelopers #cheerstothenewyear

Happy New Year! 🎈🎊🎆 We hope this new year brings you joy, growth, and elevation. The ClearSet.AI team is elated to see what this new year has in stored for us and our supporters! Cheers! 🥂🥳🤖🦾 #2025 #newyear #artificialintelligence #happynewyear #aisoftwaredevelopment

Once again, congratulations to ClearSet.AI and to other successful business owners who were apart of the USF Diversity and Skanska 2024 USF Mentor Protégé Program! We are still in awe of achieving this opportunity. 🤖🚧 “Thank you to our supporters. Thank you USF and Skanska. Thank you to my team. I am very grateful and blessed!” - Founder & CEO at ClearSet.AI 🦾💚💙

Thank you to our supporters! We hope you enjoyed your Christmas! 🎁🎄 Happy holidays from the ClearSet.AI team 🤖🦾💚 We are so grateful for the ongoing support and cherish each and every one of you. 👏🏿🤖 #happyholidays #MerryChristmas #artificialintelligence #thankyou

Merry Christmas from the ClearSet.AI team!🎄🎁 We wish you warmth, joy, and love ❤️💚🎁

Thank you FaithTech Tampa Bay chapter for having our founder Yasmine Gardiner as a panelist this Thursday 11/21 at 6pm at Embarc Collective! We are excited to be apart of this fantastic organization as a proud believer in Christ. It is He who gets the glory, and we are blessed to be a vessel. 🙏🏿 Who is FaithTech? FaithTech is a global tech community for Christ. We exist to help people in the tech ecosystem find community and steward their skills to glorify God. ✝️ #speaker #panelist #artificialintelligence #founder #ceo

ClearSet.AI presents our newest generative AI case study... Project "REDS" 🤖🏠 Chapter 5: GAN Image 1! 🏠 Let's recap chapter 4! 😀⬇ In chapter 4, we explained what preprocessing looks like in a programming format. We were able to preprocess the datasets to start training our generative adversarial network (GAN)! Now let's discuss what our GAN outputted for us: We successfully trained the GAN and obtained some intriguing results! In the image, the top row displays the input images, while the bottom row shows the generated ones. The model appears to generalize effectively! 👏 The segmented images produced can be utilized for future applications, such as feature extraction or as inputs to create realistic impressions in Unreal Engine or Unity. The GAN processes RGB images of facades and generates segmentations that can be used for realistic depth in Unity or Unreal Engine, enabling future artistic expressions. Our next step is to train on StyleGAN2, which will allow us to generate RGB images of facades. We've also developed some preliminary code in the Pix2Pix GAN Collab Notebook that stitches together four sides to create a fully segmented 3D building, ready to be imported into Unity or Unreal Engine. We want the community of Tampa Bay and its region to follow along with us on this very cool GenAI case study! 🙂 This is chapter 5 of ClearSet.AI Project REDS! 📖 Follow along the journey, as each week will be a new chapter in our case study. This is a case study dedicated to strengthening our community, environment, and construction professionals! 😃 #tampabay #genai #pinellascounty #hillsboroughcounty #stpete #clearwater #tampafl #community #realestate #realestatedeveloper #construction #casestudy #aicasestudy #tech #ai #environment #simulation #preprocessing #datasets #gans

ClearSet.AI presents our newest generative AI case study... Project "REDS" 🤖🏠 Chapter 4: Behind the "Preprocessing" Scenes 🏠 Let's recap chapter 3! 😀⬇ In chapter 3, we completed preprocessing the architectural datasets (which means adjusting, filtering, or adding to the data before analyzing it, which is an important part of cleaning "wonky" data for precise analysis). Chapter 4! 🤖⬇ Woohoo, we made it to chapter 4. 🙌 Chapter 4 is about understanding what preprocessing looks like in a programming format. Take a look at the video. The video entails how we were able to preprocess the datasets to start training our generative adversarial network (GAN). Let's discuss the video: ▶ Preprocessing in the context of the Delaunay triangulation typically involves preparing your data before applying the Delaunay algorithm (a method for triangulating a set of points such that no point lies inside the circumcircle of any triangle, resulting in a network of well-formed triangles). For geographic information systems (GIS), the Delaunay triangulation is employed to create terrain models and contour maps from elevation data. ▶ We downloaded the SVGtoPNG application to use Java code to read the SVGs into PNGs. ▶ The background of the blueprints will create the foundation for analyzing different scenarios. ▶ The doors by floorplans, parking lots, and footprints in each room by floorplans are part of strengthening the foundation of the GAN. ▶ The GAN will contain xml, png, and gom files to give full in-depth characteristics of the floor plans. ▶ The "combined_points" contained files of landmarks, districts, and historical centers in Pinellas County. ▶ The "combined_polygons" contained files of zoning areas, EMS grids, and areas of sewer and water in Pinellas County. ▶ The conditional GAN should reference the different lines and points to correctly and successfully generate an output. We are now ready to train our GANs! 👏🤖  We want the community of Tampa Bay and its region to follow along with us on this very cool GenAI case study! 🙂 This is chapter 4 of ClearSet.AI Project REDS! 📖 Follow along the journey, as each week will be a new chapter in our case study. This is a case study dedicated to strengthening our community, environment, and construction professionals! 😃 #tampabay #genai #pinellascounty #hillsboroughcounty #stpete #clearwater #tampafl #community #realestate #realestatedeveloper #construction #casestudy #aicasestudy #tech #ai #environment #simulation #preprocessing #datasets #gans 

ClearSet.AI presents our newest generative AI case study... Project "REDS" 🤖🏠 Chapter 4: Behind the "Preprocessing" Scenes 🏠 Let's recap chapter 3! 😀⬇ In chapter 3, we completed preprocessing the architectural datasets (which means adjusting, filtering, or adding to the data before analyzing it, which is an important part of cleaning "wonky" data for precise analysis). Chapter 4! 🤖⬇ Woohoo, we made it to chapter 4. 🙌 Chapter 4 is about understanding what preprocessing looks like in a programming format. Take a look at the video. The video entails how we were able to preprocess the datasets to start training our generative adversarial network (GAN). Let's discuss the video: ▶ Preprocessing in the context of the Delaunay triangulation typically involves preparing your data before applying the Delaunay algorithm (a method for triangulating a set of points such that no point lies inside the circumcircle of any triangle, resulting in a network of well-formed triangles). For geographic information systems (GIS), the Delaunay triangulation is employed to create terrain models and contour maps from elevation data. ▶ We downloaded the SVGtoPNG application to use Java code to read the SVGs into PNGs. ▶ The background of the blueprints will create the foundation for analyzing different scenarios. ▶ The doors by floorplans, parking lots, and footprints in each room by floorplans are part of strengthening the foundation of the GAN. ▶ The GAN will contain xml, png, and gom files to give full in-depth characteristics of the floor plans. ▶ The "combined_points" contained files of landmarks, districts, and historical centers in Pinellas County. ▶ The "combined_polygons" contained files of zoning areas, EMS grids, and areas of sewer and water in Pinellas County. ▶ The conditional GAN should reference the different lines and points to correctly and successfully generate an output. We are now ready to train our GANs! 👏🤖  We want the community of Tampa Bay and its region to follow along with us on this very cool GenAI case study! 🙂 This is chapter 4 of ClearSet.AI Project REDS! 📖 Follow along the journey, as each week will be a new chapter in our case study. This is a case study dedicated to strengthening our community, environment, and construction professionals! 😃 #tampabay #genai #pinellascounty #hillsboroughcounty #stpete #clearwater #tampafl #community #realestate #realestatedeveloper #construction #casestudy #aicasestudy #tech #ai #environment #simulation #preprocessing #datasets #gans 

ClearSet.AI presents our newest generative AI case study... Project "REDS" 🤖🏠 Chapter 3: Correcting Corrupt Architectural Datasets 🏠 Let's recap chapter 2! 😀⬇ In chapter two, we collaborated on which GAN (generative adversarial network) would be best for training each dataset! 😎😲 We have selected the following GANs: 1. Stylegan2 2. Pix2Pix & Pix2PixHD 3. SpadeGAN 4. CycleGAN 5. Conditional GAN (cGAN). Chapter 3! 🤖⬇ In chapter 3, we completed preprocessing, (which involves the manipulation, filtration, or augmentation of data before analysis, often serving as a crucial step in the data cleaning process) the architectural datasets, and here are our findings: Preprocessing Steps:  1 & 2: Architectural styles & Architectural Styles Periods Dataset + Resized images to 256x256 pixels. + Ensured all images are in RGB format. 3. Facades Dataset: + Ensured all images were in RGB format. + Organized images into training and testing subsets (trainA, trainB, testA, testB). 4. CubiCasa5k Dataset: + Split data into training, validation, and testing sets based on the provided text files. + Saved preprocessed images to their respective directories. 5. Structural FloorPlans: + Modified the SVG2PNG.java script in SVGtoPNG application + Modified the SVGReader.java in SVGtoPNG application 6. Floorplans Delandre Dataset + Resized to 256x256 pixels and converted.tif files to.pngs. + Split the dataset into training and testing sets. 7. ADE20K Dataset + Organized preprocessed images into appropriate training and validation datasets. 8. Mapillary Images (collected via API) + Created one Python script to collect images from Mapillary via API. + Images and the associated geolocation metadata were downloaded from API. + Extracted and saved georeferencing data alongside images in the preprocessing Python script. 9. Pinellas County Spatial Data (SHP files) + Combine multiple SHP files from the Pinellas County website.  + Saved each geometry type to separate shapefiles for future use in the conditional GAN model. 10. Pinellas County Ordinances (xlsx file) + Loaded and preprocessed text data from the Excel file. + Saved the vectors in a compressed format to ensure compatibility and efficient storage. All preprocessing tasks and script modifications have been completed and the datasets are now ready to be used to train their respective specialized GANs. 💪 We want the community of Tampa Bay and its region to follow along with us on this very cool GenAI case study! 🙂 This is chapter 3 of ClearSet.AI Project REDS! 📖 Follow along the journey, as each week will be a new chapter in our case study. This is a case study dedicated to strengthening our community, environment, and construction professionals! 😃 #tampabay #genai #pinellascounty #hillsboroughcounty #stpete #clearwater #tampafl #community #realestate #realestatedeveloper #construction #casestudy #aicasestudy #tech #ai #environment #simulation #students #learnaboutai #aitopics #aiinrealestate